Thread Subject: Re: 3d/3e for content format

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: Sean Hayes
Date: Tue, Sep 18 2007 6:30 AM

The problem here is in the framing (assuming we are going to keep it) the problem is that the phrase

"When a content format supports multiple languages..."

Is not discriminating enough, as 'support is so open ended. You can put English and Spanish in the same ASCII document fairly readily, but you couldn't reliably extract the Spanish words without a fair degree of intelligence.

Perhaps if we change the discrimination to:

"When a content format supports structured text..."

As a generalisation of 'markup' we then have a place to hang language changes of passages.

Sean Hayes
Incubation Lab
Accessibility Business Unit
Microsoft

Office: +44 118 909 5867,
Mobile: +44 7875 091385


-----Original Message-----
From: = EMAIL ADDRESS REMOVED = [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Hoffman, Allen
Sent: 17 September 2007 20:38
To: TEITAC Web/Software Subcommittee
Subject: Re: [teitac-websoftware] 3d/3e for content format

Hmmmm:

I'd say more that:

ASCII shall not be used for documents that contain tables, columns, or
other format that requires additional mark up. the 3d/3e (they moved in
the 914 release), may be problematic.

One could always use <language> as a convention in ASCII documents, but
it really wouldn't easily drive AT, of course unless AT made use of the
convention in some way--but I don't think plain-text can really
accomplish that.

The simple truth here is that plain ASCII can't do the same things text
with mark up can in terms of accessibility.

I think that is the main point to make, rather than prohibiting ASCII as
the interpretation. You may be correct that some will interpret this in
that way--we might consider an exception for the 3d for plain-text only,
but not for 3e, as language changing should require better underlying
infrastructure to support accessibility.




Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303

-----Original Message-----
From: = EMAIL ADDRESS REMOVED =
[mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Peter
Korn
Sent: Monday, September 17, 2007 3:31 PM
To: TEITAC Web/Software Subcommittee
Subject: Re: [teitac-websoftware] 3d/3e for content format

Hi Allen,

I understand your logic, and I agree with your conclusions. But I
remain concerned that we are essentially saying that governments shall
not use ASCII for documents. We have not yet settled the question of
where these provisions should be applied - just to published web
documents? to formal memos in e-mail from management? to all documents
exchanged within the government on any topic? - but when we do, I think
we will find a lot of opposition to saying that ASCII cannot be used for
that purpose.


Regards,

Peter Korn
Accessibility Architect,
Sun Microsystems, Inc.

> Peter:
>
> My take is, ASCII would not allow language identification, so
> application of that standard for a product sing that format while
> feasible will yield a noncompliant result.
>
> So lets say I'm comparing three products for selection:
>
> 1. product delivering information in ASCII in multiple languages.
> 2. Product delivering information in multiple languages in HTML.
> 3. Product delivering information in one language in ASCII.
>
> Product 1 would fail 3d and 3e.
> Product 2 would be able to pass both if coded by the vendor correctly.
> Product 3 would fail both.
>
>
> We might just write that the natural language of each portion of
> content must be programmatically determinable from mark up. wouldn't
> that encapsulate this better?
>
>
>
> Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303
>
> -----Original Message-----
> From: = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Peter

> Korn
> Sent: Friday, September 14, 2007 2:46 AM
> To: TEITAC Web/Software Subcommittee
> Subject: Re: [teitac-websoftware] 3d/3e for content format
>
> Hi Allen,
>
> I think your test is too simplistic. Do we truly intend to say that
> an ASCII document (or e-mail or...) fails? Or should we find language

> to make it clear we are talking about a rich text document format.
>
> Certainly it should be understood that switching languages in ASCII,
> there is no way for an assistive technology to know what is going on.
>
> Maybe this is more of a communication issue than a procurement
issue...
>
>
> Regards,
>
> Peter Korn
> Accessibility Architect,
> Sun Microsystems, Inc.
>
>
>> Andrew:
>>
>> It would be work to verify to a very high degree, but I don't think
>> this is an overwhelming requirement to test to a reasonable degree
>> via
>>
>
>
>> automated means.
>>
>> General test logic:
>> if content has no language identifier, it fails.
>> if it has (1) language identifier, it passes unless:
>> by character, or word analysis, it can be determined that
>>
> content
>
>> contains more than the initial language identifier.
>>
>> nailing down a specific list of failures is more difficult, but spell

>> check can probably address such a requirement by including language
>> identification as part of that function. I'd expect language
>> identification would have to be part of any multi-lingual authoring
>> tool anyway at some point.
>>
>>
>>
>>
>>
>> Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303
>>
>> -----Original Message-----
>> From: = EMAIL ADDRESS REMOVED =
>> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of
>> Andrew Kirkpatrick
>> Sent: Thursday, September 13, 2007 9:15 AM
>> To: TEITAC Web/Software Subcommittee
>> Subject: Re: [teitac-websoftware] 3d/3e for content format
>>
>> Just for the record, my objection is more to the "changes in
language"
>> section. Seems like a lot of extra work to verify this...
>> AWK
>>
>>
>>
>>> -----Original Message-----
>>> From: = EMAIL ADDRESS REMOVED =
>>> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of
>>> Hoffman, Allen
>>> Sent: Wednesday, September 12, 2007 2:34 PM
>>> To: TEITAC Web/Software Subcommittee
>>> Subject: Re: [teitac-websoftware] 3d/3e for content format
>>>
>>> is there synthetic speech for Creole? Just curious!
>>>
>>>
>>>
>>>
>>> Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303
>>>
>>> -----Original Message-----
>>> From: = EMAIL ADDRESS REMOVED =
>>> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of
>>> Smith, Jamie
>>> Sent: Wednesday, September 12, 2007 2:29 PM
>>> To: TEITAC Web/Software Subcommittee
>>> Subject: Re: [teitac-websoftware] 3d/3e for content format
>>>
>>> I definitely feel we must have this as a provision. In Florida,
>>> public information is often in Spanish and Creole as well as
English.
>>>
>>>
>>
>>
>>> Some documents are the "Florida Comprehensive Assessment Test" test
>>> materials and while the document is in Spanish or Creole the test
>>> examples are always in English. The lack of coding for the language

>>> did cause speech to select a language and attempt to read the Creole

>>> with that language.
>>> >From the point of the first change to Creole on (even when text was
>>>
>>>
>>>> back
>>>>
>>>>
>>> to English), speech messed up. Adding the language code in the
>>> doctype and then to the specific sections which had a different
>>> language fixed the problem. The reason the language code wasn't
>>> done,
>>>
>>>
>>
>>
>>> simply because it wasn't required.
>>>
>>> -----Original Message-----
>>> From: = EMAIL ADDRESS REMOVED =
>>> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of
>>> Andrew Kirkpatrick
>>> Sent: Thursday, September 06, 2007 9:32 AM
>>> To: TEITAC Web/Software Subcommittee
>>> Subject: Re: [teitac-websoftware] 3d/3e for content format
>>>
>>> In the spirit of trying to reduce the extra weight of the new
>>> standards, I'll raise my question again -- why is language support
>>> in
>>>
>
>
>>> 508? We do have other WCAG standards that we have not included, and

>>> this seems like it is not needed in the U.S. standard.
>>>
>>> What is our rationale for requiring the language support in content
>>> formats and web and software provisions?
>>>
>>> AWK
>>>
>>>
>>>
>>>> -----Original Message-----
>>>> From: = EMAIL ADDRESS REMOVED =
>>>> [mailto: = EMAIL ADDRESS REMOVED = ] On
>>>>
>>>>
>>> Behalf Of Peter
>>>
>>>
>>>
>>>> Korn
>>>> Sent: Tuesday, September 04, 2007 3:56 PM
>>>> To: TEITAC Web/Software Subcommittee
>>>> Subject: Re: [teitac-websoftware] 3d/3e for content format
>>>>
>>>> Hi Allen,
>>>>
>>>> We need to be careful with this one. Simple text editors (think
>>>> Notepad) won't be able to do this.
>>>>
>>>>
>>>> Regards,
>>>>
>>>> Peter Korn
>>>> Accessibility Architect,
>>>> Sun Microsystems, Inc.
>>>>
>>>>
>>>>
>>>>> Action item from Tuesday, 09/04.
>>>>>
>>>>> Content formats which support multiple languages MUST provide a
>>>>> programmatically determinable mechanism to identify the primary
>>>>> language, and the language of any sections which are in
>>>>>
>>>>>
>>>> another language
>>>>
>>>>
>>>>> from the primary language.
>>>>>
>>>>>
>>>>>
>>>>> Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303
>>>>>

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