Re: [Maturity] Common HTML Entities

I'm seeing a few questions here that I'll try to pull apart, in hopes of 
answering the right one:

  * W3C recommendations about web content validation are mostly the
    charge of the HTML / WhatWG group. We don't want to touch that space.
  * W3C requirements for its own formal publications are that they pass
    the W3C validator. For several years now, that is using the
    validator.nu codebase and it is actively maintained. It does
    validate to a more permissive syntax than the previous W3C validator
    did.
  * There are no formal requirements on our validation practices for
    editors' drafts, as long as they can produce documents that meet the
    requirements. We are using Respec in the browser, meaning as long as
    the browser can produce a DOM, Respec should be able to output
    something.
  * In spite of all this, I continue to find it valuable to use
    well-formed XML, known as polyglot HTML, in my document sources. The
    stricter validation can find authoring errors that looser validation
    won't, and I have seen instances of those errors leading to missing
    content. It is very hard to notice it missing in the rendered view,
    and it can change intended meaning. Therefore I strongly recommend
    adopting this practice.

Michael

On 2022-08-23 7:21 a.m., Janina Sajka wrote:
> Thanks for looking back through Maturity Model emails by way of coming
> up to speed in the subgroup, Lionel.
>
> The context of my email was a conversation on marking up the document
> which would be forwarded for First Public Working Draft publication. It
> was not meant as guidance for the wider world of web publishing. When
> APA inherited Maturity Model work from AGWG, the document under
> development was hosted in Google Docs. Mary Jo took on the task of
> primary role of recoding into html, as that's what is needed to publish.
>
> PS: The validatornu package that started out in W3C is generally
> available across Linux platforms, and probably on other platforms as
> well. I can't speak to the level of maintanance it receives or might
> need.
>
> Name            : validatornu
> Version         : 20.6.30-1
> Description     : The Nu Html Checker, a command line tool for HTML5
> markup validation
> Architecture    : any
> URL             :https://validator.github.io/validator/
>
> Best,
>
> Janina
>
> Lionel Wolberger writes:
>> I didn't want this to get lost, so I am adding my two cents.
>>
>> First, there is the validity of the HTML itself.
>>
>> Currently, HTML can include anything. It is barely validated at all. I
>> understand this is the result of an industry-wide consensus, due to the
>> limitations on authoring and the exigencies forced upon browsers operating
>> in the wild.
>>
>> At the other extreme of this zero to 100 conformance scale is 100% valid
>> HTML as per the W3C Validator. In my experience this has some use in
>> certain settings, but in general has been sidelined. People first check how
>> Google parses their pages, and often do not validate at all. I hear that
>> updating this validator is also resource-intensive.
>>
>> Second, there is the semantic coherence of the HTML. The trend of
>> frameworks to write HTML in endlessly nested DIVs and SPANs, and "low code"
>> solutions to distance the developer from the HTML, have contributed to the
>> proliferation of non-semantic HTML, and the associated proliferation of
>> ARIA roles and labels.
>>
>> So your proposal, Janina, seems to me to be proposing a validation
>> somewhere between the two HTML validation extremes. Not zero, and not 100%.
>> And it proposes a new type of validation, one that would assess a heap of
>> DIVs and ARIA labels as poorer than a Button.
>>
>> Are you proposing a wiki page, "Common HTML entities used by Screen
>> Readers"?
>>
>> Would that meet the need?
>>
>> HTH
>>
>> - Lionel
>>
>>
>>
>>
>>
>>
>> [image: Lionel Wolberger]
>> <https://userway.org/?utm_source=email-signature&utm_medium=email&utm_campaign=Lionel_Wolberger>
>>
>> Lionel Wolberger
>> COO, UserWay Inc.
>> [image: UserWay.org]
>> <https://userway.org/?utm_source=email-signature&utm_medium=email&utm_campaign=Lionel_Wolberger>
>> [image: Phone Icon] +1 (415) 800-4557 <+14158004557>
>> [image: Envelope Icon]
>> lionel@userway.org  <lionel@userway.org>
>> [image: Schedule with Me]
>> <https://userway.org/s/lionel?utm_source=email-signature&utm_medium=email&utm_campaign=Lionel_Wolberger>
>>
>>
>>
>>
>> On Wed, May 18, 2022 at 10:17 PM Janina Sajka<janina@rednote.net>  wrote:
>>
>>> Colleagues:
>>>
>>> During today's Maturity Model teleconference we noted the value of
>>> cleanly marked up HTML. I even mentioned APA Co-Chairs are considering
>>> whether we might want to design and adopt a cross-platform tool that
>>> would help us enforce certain styling--which is why I've cc'd this email
>>> to the APA Chairs list!
>>>
>>> I mentioned I have a bookmark that helps me when I'm editing. It's
>>> actually a W3C resource available here:
>>>
>>> https://www.w3.org/wiki/Common_HTML_entities_used_for_typography
>>>
>>>
>>> Note that this does not reference elements such as <q> and </q> for
>>> "smart" quotation marks.
>>>
>>> I would note further from my personal experience that sometimes simple
>>> CSS rules can help. In recent years I used CSS to mark quoted text in
>>> dark red in order to keep my colleagues from trying to edit our
>>> quotations! Yes, we can quote, quote less, or quote more; but we can't
>>> edit what we quote from another publication. One would think that would
>>> be obvious, so I tried to make it obvious where those quotes were.
>>>
>>> hth!
>>>
>>> Janina
>>>
>>>
>>> --
>>>
>>> Janina Sajka (she/her/hers)
>>> https://linkedin.com/in/jsajka
>>>
>>> The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
>>> Co-Chair, Accessible Platform Architectureshttp://www.w3.org/wai/apa
>>>
>>>
>>>