Re: "image analysis heuristics" (ISSUE-66)

Maciej Stachowiak wrote:
> On Feb 5, 2010, at 5:09 AM, Matt May wrote:
>> Secondly, after all this discussion on the list about this
>> particular sentence, and after several of us (including yourself,
>> Maciej, and Lachlan) came to some kind of agreement on what should
>> be said (and more importantly, where it should be said), it's
>> especially galling to see Ian unilaterally making a minor change
>> that clearly shows his disregard for the WG's input.
>
> I don't remember getting down to the level of a specific single
> wording, though we did come up with a number of suggestions. That
> being said, I'm more interested in hearing what further changes would
> lead to a change you are happy with. Do you have specific specific
> requests? (Also happy to hear from Lachlan or anyone else who has an
> opinion on the matter.)

I'm not concered about the specific wording used, I'm generally happy 
for editors to phrase things in their own way.  I think what might be 
most productive here is that, instead of us trying to lock down and 
agree on specific wording, we should come up with a list of objective 
criteria to evaluate the any text Hixie writes in the spec related to 
this issue.

Since there appears to have been a general agreement about what I 
previously suggested, excluding the list of techniques, it seems useful 
to compare my suggestion with the current spec text.  For reference, 
here is what I suggested earlier, minus the list of techniques:

---
When the user is unable to make direct use of the image, e.g. due to a
visual disability or because they are using a text terminal with no
graphics capabilities, user agents may also provide the user with the
ability to obtain any other information about the image that may assist
the user in understanding its content or purpose, utilising any
available repair technique.

For more information, refer to the User Agent Accessibility Guidelines
techniques for repairing missing content ([UAAG10-TECHS], section 2.7).
---

And here is the current spec text:
---
User agents may also apply heuristics to help the user make use of the 
image when the user is unable to see it, e.g. due to a visual disability 
or because they are using a text terminal with no graphics capabilities. 
Such heuristics could include, for instance, optical character 
recognition (OCR) of text found within the image.

Warning! While user agents are encouraged to repair cases of missing alt 
attributes, authors must not rely on such behaviour. Requirements for 
providing text to act as an alternative for images are described in 
detail below.
---

I think the warning to authors about not relying on repair techniques is 
non-objectionable, and may be left as is.

Based on the discussion, I've come up with the following proposed criteria:

1. Permits user agents to apply any repair technique
2. Does not list specific techniques.
3. Informatively references the techiques described in UAAG (either 1.0
    or the 2.0 draft)
4. The text is located in a section dedicated to describing
    implementation requirements and is clearly distinguished from
    authoring requirements.
5. Does not imply the use of futuristic technologies.
6. Does not imply that any technique can reliably determine author
    intent.
7. Indicates that it is about providing additional information about the
    image, which may help the user to understand the image's content or
    purpose.

(If I've missed any, or included any that I shouldn't have, feel free to 
comment or make adjustments)

To understand what exactly is wrong with the current spec text, it would 
be useful for people who are objecting to it, to clearly describe how it 
fails to meet any of these criteria.

In my own view, the current text fails #2 by mentioning OCR, #3 by not 
referencing UAAG; partially fails #5 by mentioning "heuristics" without 
clearly describing what would or would not be classified as such, and 
also fails #7.

I also believe the location of the text in the current spec is 
acceptable, as it is in the section describing implementation 
requirements for images.  This is consistent with where implementation 
requirements are placed for every other element in the spec and should 
not be treated any differently.  I strongly disagree with Matt about 
moving this text to the Web Browsers section, where it has no relevance 
to anything else in that section.

-- 
Lachlan Hunt - Opera Software
http://lachy.id.au/
http://www.opera.com/

Received on Saturday, 6 February 2010 10:53:23 UTC