Re: Working Group Decision on ISSUE-31 / ISSUE-80 validation survey

Maciej, and the other chairs,

	I have 3 prompts for clarification:

1) Does this letter and in my replies to Aryeh [*], in combination with 
the info in the formal objection [#] bring new info to the table, 
according to the chairs evaluation? Do I need to list the specific 
points that I think are new?

2) Please confirm that I have perceived the following correctly: The 
generator exception says that lack of alt should not be reported when 
generator is present. In another part of the decision you say that <img 
src=* role=presentation> is forbidden. Thus, <img src=* 
role=presentation> is forbidden except when there is meta generator 
string. (If so, then there actually *is* a form of per element 
generator exception.) Likewise: HTML5 says that when IMG is the sole 
content of an A element, then there, quote, "must" be alt text. Except 
when there is a generator string.

3) Please clarify what it means when HTML5 says that pages with the 
generator string are not conforming: 

]]This case does not represent a case where the document is conforming[[

   While the Decision says that they _are_ conforming:

]]The presence of <meta name=generator> makes missing alt conforming[[

	End of prompts for clarification.

Statement of new fact/opinion: 

	It seems that if the validator only performs a second degree 
validation when 'generator' string is present, then this fact should 
not be kept hidden from the persons which are making use of the 
validator, be they users, authors, someones responsible for quality 
control - as well as generator developers themselves. For the generator 
developers, this exception means that if the author wants to get full 
quality validation, then he/she has to remove the generator string - 
thus loosing the generator info, and with possible worse ranking on 
stats etc as a result. Identification is usually a matter of pride, but 
this decision connect the generator string with lower quality. The 
irony is that even HTML Tidy spits out a generator string ...  
Validators as authoring tools fail to be authoring tools if authors 
cannot get back relevant info from them without editing their code 
first (that is: removing the generator string).

	For authors and tool vendors, the visibility of the fact the page is 
validated according to a lower bar is a matter of fair competition and 
same rules. Some kind of "consumer html", which the big tool vendors 
can spit out, without getting any negative validation messages of any 
kind, is unfair competition. We could have "loose" or "transitional" 
stamp somewhere, that *anyone* could use. But this, which some 
completely undefined group of "hand-authored" pages are barred from 
using, is not anything we can have.
leif halvard silli

Received on Tuesday, 19 April 2011 07:21:27 UTC