[whatwg] Conformance for Mail clients (and maybe other WYSIWYG editors)

[Maciej Stachowiak]

Hi Karl,

On Apr 15, 2007, at 11:48 PM, Karl Dubost wrote:

>
> in a drag and drop scenario in your mail.app or other HTML  
> authoring tool, you could imagine:
>
>  +------------------+
>  |                  |
>  |                  |  the image itself
>  |                  |
>  |                  |
>  |                  |
>  +------------------+
>  |                  |  <- here a dynamic text area popping up
>  +------------------+     to edit the content.
>
> When the image is put in the window, a text is requested by the UI  
> (a bit ala ajaxy flickr.)
> Then the markup could be generated.

I don't think this is workable for HTML-generating mail clients or  
other tools for non-experts. You paste in an image and an area pops  
up to edit text that won't actually be visible in the mail message.  
This is likely to cause confusion and annoyance to users, and pretty  
unlikely to lead to good quality alt text. People would either just  
press enter, or type a description or caption (and then be annoyed  
when it disappeared) instead of substitute text. Consider in  
particular the case where you paste in a chunk of rich text content  
containing multiple images.

Furthermore, alt text has a specific purpose, which is to make  
content with images more meaningful on non-visual or strictly text- 
only media, or for people with special accessibility needs, or for  
indexing by data mining UAs like search engines. But when I shoot off  
an email to my normally sighted GUI-using friend Bob, I know that  
none of these really apply. So including alt text for the super-funny  
kitten picture I just found on the net doesn't actually give any of  
the benefits for which alt is intended.

Now, suppose I am writing to my visually impaired friend Mary who  
uses a screen reader. In that case, I probably wouldn't send her an  
image at all. And if I were sending mail to both Mary and Bob, I  
might include an image and a visible description of it in the prose.  
That's a more comprehensible model for users than alt text. But it  
doesn't make an image like this semantically void.

Flickr is again pretty analogous here. It gives me dynamic text boxes  
next to the image, but they edit the title and the caption. Adding  
another one to edit the alt text would only confuse people, and would  
be unlikely to lead to quality alt text.

I think it remains the case that for end-user generated content,  
there will often be semantically meaningful images that are  
meaningful in themselves and cannot be considered alternate  
representations of some piece of text.

Regards,
Maciej